| __init__.py |
|
0 |
| android |
|
|
| apple |
|
|
| audio_quality |
|
|
| autoroller |
|
|
| binary_version_check.py |
|
1252 |
| BUILD.gn |
|
758 |
| chromiumos |
|
|
| clang_tidy.py |
Invoke clang-tidy tool.
Usage: clang_tidy.py file.cc [clang-tidy-args...]
Just a proof of concept!
We use an embedded clang-tidy whose version doesn't match clang++.
|
3395 |
| configure_pipewire.py |
This script is a wrapper that loads "pipewire" library.
|
2313 |
| coverage |
|
|
| cpu |
|
|
| download_tools.py |
Downloads precompiled tools.
These are checked into the repository as SHA-1 hashes (see *.sha1 files in
subdirectories). Note that chrome-webrtc-resources is a Google-internal bucket,
so please download and compile these tools manually if this script fails.
|
1883 |
| ensure_webcam_is_running.py |
Checks if a virtual webcam is running and starts it if not.
Returns a non-zero return code if the webcam could not be started.
Prerequisites:
* The Python interpreter must have the psutil package installed.
* Windows: a scheduled task named 'ManyCam' must exist and be configured to
launch ManyCam preconfigured to auto-play the test clip.
* Mac: ManyCam must be installed in the default location and be preconfigured
to auto-play the test clip.
* Linux: Not implemented
NOTICE: When running this script as a buildbot step, make sure to set
usePTY=False for the build step when adding it, or the subprocess will die as
soon the step has executed.
If any command line arguments are passed to the script, it is executed as a
command in a subprocess.
|
3224 |
| executable_host_build.py |
This script builds a GN executable targeting the host machine.
It is useful, for example, for mobile devices performance testing where
it makes sense to build WebRTC for a mobile platform (e.g. Android) but
part of the test is performed on the host machine (e.g. running an
executable to analyze a video downloaded from a device).
The script has only one (mandatory) option: --executable_name, which is
the output name of the GN executable. For example, if you have the
following executable in your out folder:
out/Debug/random_exec
You will be able to compile the same executable targeting your host machine
by running:
$ vpython3 tools_webrtc/executable_host_build.py --executable_name random_exec
The generated executable will have the same name as the input executable with
suffix '_host'.
This script should not be used standalone but from GN, through an action:
action("random_exec_host") {
script = "//tools_webrtc/executable_host_build.py"
outputs = [
"${root_out_dir}/random_exec_host",
]
args = [
"--executable_name",
"random_exec",
]
}
The executable for the host machine will be generated in the GN out directory
(e.g. out/Debug in the previous example).
|
3118 |
| get_landmines.py |
This file emits the list of reasons why a particular build needs to be clobbered
(or a list of 'landmines').
|
3235 |
| gn_check_autofix.py |
This tool tries to fix (some) errors reported by `gn gen --check` or
`gn check`.
It will run `mb gen` in a temporary directory and it is really useful to
check for different configurations.
Usage:
$ vpython3 tools_webrtc/gn_check_autofix.py -m some_mater -b some_bot
or
$ vpython3 tools_webrtc/gn_check_autofix.py -c some_mb_config
|
6218 |
| gtest-parallel-wrapper.py |
This script acts as an interface between the Chromium infrastructure and
gtest-parallel, renaming options and translating environment variables into
flags. Developers should execute gtest-parallel directly.
In particular, this translates the GTEST_SHARD_INDEX and GTEST_TOTAL_SHARDS
environment variables to the --shard_index and --shard_count flags, renames
the --isolated-script-test-output flag to --dump_json_test_results,
and interprets e.g. --workers=2x as 2 workers per core.
Flags before '--' will be attempted to be understood as arguments to
gtest-parallel. If gtest-parallel doesn't recognize the flag or the flag is
after '--', the flag will be passed on to the test executable.
--isolated-script-test-perf-output is renamed to
--isolated_script_test_perf_output. The Android test runner needs the flag to
be in the former form, but our tests require the latter, so this is the only
place we can do it.
If the --store-test-artifacts flag is set, an --output_dir must be also
specified.
The test artifacts will then be stored in a 'test_artifacts' subdirectory of the
output dir, and will be compressed into a zip file once the test finishes
executing.
This is useful when running the tests in swarming, since the output directory
is not known beforehand.
For example:
gtest-parallel-wrapper.py some_test \
--some_flag=some_value \
--another_flag \
--output_dir=SOME_OUTPUT_DIR \
--store-test-artifacts
--isolated-script-test-output=SOME_DIR \
--isolated-script-test-perf-output=SOME_OTHER_DIR \
-- \
--foo=bar \
--baz
Will be converted into:
vpython3 gtest-parallel \
--shard_index 0 \
--shard_count 1 \
--output_dir=SOME_OUTPUT_DIR \
--dump_json_test_results=SOME_DIR \
some_test \
-- \
--test_artifacts_dir=SOME_OUTPUT_DIR/test_artifacts \
--some_flag=some_value \
--another_flag \
--isolated-script-test-perf-output=SOME_OTHER_DIR \
--foo=bar \
--baz
|
8206 |
| gtest_parallel_wrapper_test.py |
|
6915 |
| ios |
|
|
| iwyu |
|
|
| libs |
|
|
| matlab |
|
|
| mb |
|
|
| msan |
|
|
| network_emulator |
|
|
| OWNERS |
|
79 |
| perf |
|
|
| presubmit_checks_lib |
|
|
| remove_extra_namespace.py |
Remove extra namespace qualifications
Looks for names that don't need to be qualified by namespace, and deletes
the qualifier.
Depends on namespace names being properly formatted
|
3415 |
| sanitizers |
|
|
| sslroots |
|
|
| ubsan |
|
|
| version_updater |
|
|
| video_quality_toolchain |
|
|
| vim |
|
|
| whitespace.txt |
|
667 |